Almost all older adults experience cognitive frailty with age, with around one-third of adults aged 85 or older suffering from Alzheimer?s disease (AD). Cognitive frailty, particularly Alzheimer?s disease, threatens to overwhelm medical resources in the United States and much of the developed world. Therefore, it is important that we learn how to optimize and maintain cognitive performance in cognitively-frail older adults, particularly those who are at high-risk of suffering from AD, such as adults over 70 years of age where more than half express AD pathology. The present multi-arm randomized control trial takes a novel theory-driven approach to enhancing cognition in older adults by training them to flexibly deploy attentional focus in working memory. Flexibility in allocating and switching attentional resources will be trained by having participants respond to unpredictable cues in working memory. The ability to flexibly and efficiently allocate attentional control underlies successful performance on a broad array of cognitive tasks. Hence, training in this area may enhance performance not only on related tasks (near transfer), but also on tasks that are perceptibly not related to the training task (far transfer). The current proposal utilizes game-based simulations in all training arms in healthy older adults. The study has three training arms. The first two arms use experimenter-designed simulation games, where participants will be trained on either predictable low attentional control (Arm 1) or unpredictable high attentional control (Arm 2) working memory games. The third arm uses a commercially-available strategy video game requiring highest level of attentional control, by adding multi-tasking to the unpredictable attentional shifts in working memory. In all three training arms, neural and cognitive changes will be examined immediately after intervention; cognitive changes will also be assessed at +6mo period after training completion. Additionally, a single-session, baseline neuroimaging data will be collected in a comparison group of younger adults (no training in this group). We expect that the high attentional control training arms will greatly improve cognition in older adults, with cognitive frailty interacting with the extent to which attentional control is trained. High attentional control training arms are also expected to heighten compensatory brain activation after intervention, for both near and far in-scanner transfer tasks, mimicking the baseline activity of younger brains. These training arms are also expected to positively impact brain structures that progressively decline with aging. This clinical trial will result in development of behavioral intervention tools, which will have the potential to delay the onset of memory-related disorders, such as AD, by instantiating durable improvements in cognitive functions in older adults. Such interventions can not only improve an individuals? quality of life, but also decrease the financial burden of a rapidly aging society.